2 files changed, 27 insertions, 0 deletions
diff --git a/src/htscharset.c b/src/htscharset.c
index 6386c09..614f8b4 100644
--- a/src/htscharset.c
+++ b/src/htscharset.c
@@ -1192,6 +1192,28 @@ hts_UCS4* hts_convertUTF8StringToUCS4(const char *s, size_t size, size_t *nChars
   return dest;
 }
 
+int hts_isStringUTF8(const char *s, size_t size) {
+  const unsigned char *const data = (const unsigned char*) s;
+  size_t i;
+
+  for(i = 0 ; i < size ; ) {
+    hts_UCS4 uc;
+
+    /* Reader: can read bytes up to j */
+#define RD ( i < size ? data[i++] : -1 )
+
+    /* Writer: upon error, return FFFD (replacement character) */
+#define WR(C) if ((C) == -1) { return 0; }
+
+    /* Read Unicode character. */
+    READ_UNICODE(RD, WR);
+#undef RD
+#undef WR
+  }
+
+  return 1;
+}
+
 char *hts_convertUCS4StringToUTF8(const hts_UCS4 *s, size_t nChars) {
   size_t i;
   char *dest = NULL;
diff --git a/src/htscharset.h b/src/htscharset.h
index 92f8b7c..0b5b4f2 100644
--- a/src/htscharset.h
+++ b/src/htscharset.h
@@ -87,6 +87,11 @@ extern char *hts_getCharsetFromMeta(const char *html, size_t size);
 extern int hts_isStringAscii(const char *s, size_t size);
 
 /**
+ * Is the given string an UTF-8 string ?
+ **/
+extern int hts_isStringUTF8(const char *s, size_t size);
+
+/**
  * Is the given charset the UTF-8 charset ?
  **/
 extern int hts_isCharsetUTF8(const char *charset);